SubTechniquepersistenceai-attack-stagingATLAS

AML.T0018.002Embed Malware

What it is

Adversaries may embed malicious code into AI Model files. AI models may be packaged as a combination of instructions and weights. Some formats such as pickle files are unsafe to deserialize because they can contain unsafe calls such as exec. Models with embedded malware may still operate as expected. It may allow them to achieve Execution, Command & Control, or Exfiltrate Data.

References

  1. https://atlas.mitre.org/techniques/AML.T0018.002

Related by meaning· 6

Nearest entities by semantic similarity across the cs-graph corpus.

ATLAS
Manipulate AI Model
ATLAS
Corrupt AI Model
ATLAS
AI Agent Tool Poisoning
ATLAS mitigation
Vulnerability Scanning
ATLAS
Machine Compromise
ATLAS
AI Agent Tool Data Poisoning
Sourced from MITRE ATLAS — Adversarial Threat Landscape for AI Systems. Curated by Adam Lundqvist, SQUR.